회사채 신용등급 예측에 관한 연구 (A Study on Forecasting Credit Rates of Corporate Bonds)

서지정보

· 발행기관 : 한국회계정보학회
· 수록지 정보 : 회계정보연구 / 21권 / 25 ~ 59페이지
· 저자명 : 김태정, 문남희, 문성주

초록

본 연구는 신용등급예측 모형 중에서 가장 널리 사용되는 판별모형을 이용하여 신용평가 3사가 IMF 외환위기 이후 신용평가의 예측력을 강화하기 위해 채택한 재무변수가 선행연구에서 그 유용성이 밝혀진 재무변수들보다 나은지를 살펴보고, 국내 신용평가 3사가 신용평가 예측력을 강화하기 위해 중시한 비재무변수가 판별모형에 추가적으로 감안될 경우 재무변수만을 가지고 추정한 모형보다 예측력이 향상되는지를 살펴보고자 하였다. 그 주요한 연구 결과는 다음과 같다. 첫째, 재무표만을 이용한 판별분석 결과 특정기간에 한정되지만 국내 신용평가 3사가 채택한 재무변수인 기업경상이익률, 단기차입금의존도 등은 신용등급예측에 유용한 지표가 될 수 있음을 알 수 있다. 즉, 기업경상이익률이 양호할수록 회사채 신용등급에 양(＋)의 영향을 미치는 반면에 단기차입금 의존도가 높을수록 회사채 신용등급에 음(－)의 영향을 미침을 알 수 있다.

영어초록

The purpose of this study is that the financial information selected by three credit rating institutions after IMF financial crisis is still valid when compared to the financial information used in previous studies. In addition, this study evaluates whether or not the prediction improves by including non-financial information in the analysis of the Discriminant model.
The results of the study are as follows. First, the study found that the ratios of operation income and the dependence on short term borrowing are useful variables among the financial information selected by the three credit rating institutions. The company with a higher rate of operation inomce received a positive credit rating for its corporate bond and the company with a higher dependence on short term borrowing received a negative credit rating for its corporate bond.
Second, when non-financial information was added to the financial information for Discriminant analysis, the study found that the company's governance structure(Large Scale Business Group) and its industry characteristic, as classified by the Korean Industrial Standard Classification , affected the credit rating of the corporate bonds. The accuracy of prediction using year 2000 financial and non-financial information in determining year 2001 corporate bond credit ratings was 79.3%. The accuracy of prediction using year 2001 financial and non-financial information in determining year 2002 corporate bond credit ratings was 78.9%. In contrast, when finance information was only used, the accuracy of prediction year 001 and 2002 credit ratings was 73.7% and 77.9%. Therefore, the study concluded that the credit ratings system that utilized not only financial information but also non-financial information would provide better credit rating predictability.
Although non-financial information alone can not improve predictability in corporate bond rating, the study illustrates that there is a significant need for further development of non-financial information to complement financial information beyond the variables used in this study.

참고자료

· 없음